I don't understand how they get their instruments so BIG!

[Christopher Burke]

Dear Anyone.

 

OK, I know they're using real instruments in the waveform example I'm going to show you and I'm using Edirol Orchestral but I don't THINK that's the point.  The point IS - they're ending up with a hewge great waveform that doesn't sound Iron Maiden loud, just full and rich, every instrument sounds full and BEEG (not overloud, just fat and round and great!) I could send you a ton of similar wave forms from synth tracks or even other Edirol Orchestral tracks that look the same. 

Mine sound mixed but the waveforms are TINY in comparison and the instruments flat don't have the fullness and fatness and rich-sounding-ness of all these other guys.  And I flat don't know what they're doing to them to get the fullness and fatness and richness.  I don't mean 'and then how do they mix them' because that's up to me to try.  (Though I might well be asking questions later!)  There's SOMETHING I don't understand that they do understand that lets them get lovely full rich sounds and keeps mine sounding tiny.  Not quiet - the track overall sounds as loud as theirs - but the individual instruments don't have the fullness of theirs, even those using the same software I am.

I'm sorry the above sounds a bit garbled, if I understood what I was talking about I'd be able to do it! It's not even quality of samples because I had a demo version of East West and hit exactly the same problem.  So it's something basic I'm not seeing. 

 

When you look at the attached wave forms, remember my mix sounds as loud as theirs as a piece - it's hitting about -4 decibels in Audacity, just like theirs does.  But look at the size and fullness difference!!

Anyone got any ideas how they get the individual instruments sounding so rich and full without red clipping the entire mix? I'll take all ideas on board and try them all out. 

[Christopher Burke]

Even though it's utterly obvious, just saying, the top one's my wave form, the hewge bar steward is theirs!!

Yours respectfully

Chris.

[haim shim]

Hi Christopher, I'm new here this is my second post.

I am also contemplating on the same question.

 

One thing that is helping me a lot lately is two things:

1. using limiter - this boots anything you want (I mainly use it on my master).

2. trying to make my mix a bit (or a lot) more leveled.. that means - I try to have as less peaks as I can, so then when I do boost, it will not sound too bad.

[john]

There are a number of contributing factors:

 

Input Gain

Had it been a mic input then “pad” buttons should also be set according to what your mic delivers

EQ,  Compressor and Limiter

 

Before you go using these treatments, you first have to work out what is going on. I suspect that it is a few peaks reaching that -4dB value. Scan the waveform to find those peaks and zoom in to eyeball the wave. When you find where the peak occurs, try  selecting some of the wave that doesn’t include the peak. Now scan to find the new peak within the selection.

 

 

 

[Mahesh]

As mentioned in the above posts, it is a number of factors that matters in regard to this.

 

 

• Recording DT (Double tracking) : One of the most common ways of getting a nice big sound is to double track your takes. ie., record two copies of the exact same thing. Because of the slightest of difference in the takes, you begin to gain some thickness and body to the part.
• Input gain levels : You want your recorded signal to be healthy but not over bearing. This is considered anywhere between -12dbB to upt -8dB based on the instrument being recorded.
• EQ - Sometimes, subtracting sound from your recorded signal to accentuate the characteristics of the instrument can result in the your track sounding bigger (that's right! relativity is wonderful).
• Compression : This is something that could really help if used right but can totally mess thing too. Because you're limiting the dynamic range of the recorded track, you start gaining some relative loudness because of it but it also makes the quietest and loudest parts close to each other so that your track sounds 'confident'.
• Stereo Width / Panning : You can quite literally make a track sound 'big' by increasing the stereo width of track using a plugin. You find a lot of free ones these days. It does have an effect of the tonal quality of the recorded track so it should be done so carefully. A good way of doing it is to apply stereo widening to a second duplicate copy of your recorded track  with heavy compression. Dial this compressed widened sound up in levels slowly and set it to your desired spot. This way you have one copy of your original track in the center and another widened copy adding some 'body' on either sides.

 

 

[Mahesh]

7 hours ago, Patchez said:

as close as possible to 0 dB

 

That's an interesting approach Patchez.

 

As mentioned earlier, I personally wouldn't go beyond - 8dB to -6dB for them gain levels given that the mixed and mastered track needs to be peaking at - 2db to "almost but not 0dB" as a standard that we hear for tracks these days. If you do record so closely to 0dB with each instrument, you'd be having to bring down each of your recorded tracks as you start mixing them together. Introducing compression just to bring it down may change the tone or quality of the recorded signal in unpredictable ways. That's additional mixing signed up for before mixing itself if you know what I mean. It may cause altercations like the mix sounding squashed etc. 

 

Moreover, with such hot levels, you don't get a lot of head room which helps to have as mentioned especially when you're trying to make things sound big spacially but also harmonically.

 

Anyways, interesting to hear the various approaches 

[john]

I think there are two fundamental concepts worth understanding here that should help address the OP's issue:

 

Headroom

The amount of signal amplitude range between the nominal signal level and the signal level at which clipping will occur. Headroom is normally expressed in dB.

 

Signal to Noise Ratio (SNR)

The measure of the level of a desired signal compared with the level of background noise. SNR is the ratio of the signal power to the noise power, usually expressed as dB.

 

In essence:

• Headroom is the room between the signal and the ceiling.
• SNR is the room between the signal and the floor.

Ideally you want your signal to be much closer to the ceiling than the floor.

 

It shouldn't be too close to the ceiling (0dB in digital)because then you run the risk that any signal treatment that results in raising the signal level, even when it just applies to certain frequencies, runs the risk of 'clipping'.

 

Clipping
Distortion due to overloading of an audio signal. In the Analogue domain this is normally caused by the peak signal voltage being limited by the audio circuits’ power supply voltage. In the Digital domain, clipping occurs at 0dB.

 

Your signal has various peaks and troughs. Any of those peaks can be the loudest point. That includes outwith the range of human hearing. Any of the troughs could be the quietest point in the signal.

 

The trouble is, the point that is the loudest can be a split second, and it can be much louder than the rest of the signal. If you were to scan your audio you could detect a peak at -1dB, but the nominal signal could be sitting around -18dB, leaving your music pretty quiet. Someone else's song could have a peak at -3dB but the rest of the song could nominally be -6dB. Their song will sound much louder (although dynamically, much flatter)

 

Dynamics

Dynamics of a piece is the variation in loudness between notes or phrases.

 

During recording, you should aim to accurately capture the performance, with a good signal level, but without any clipping (reaching 0dB).

 

Ok, so say you record some audio, and there are some large peaks while the nominal audio is pretty quiet. Turn up too loud and you increase the overall noise floor in the song. Additionally, some of those peaks might be outwith the range of hearing, so first of all you should apply a high pass EQ filter at say 40Hz on drums and bass, and at say 100Hz on a vocal track.

 

You should apply a low pass filter at at least 20 kHz. Often it can be much lower, but be wary of removing character shaping harmonics.

 

So now you might want to reduce the difference between your highest peaks and the nominal level. To do this we use a compressor.

 

You can find out about the basic controls on a compressor here:

 

https://www.songstuff.com/recording/article/compression/

 

There are a load of compression tutorials on YouTube and various audio sites.

 

In essence you look to reduce the signals above a dB threshold by the ratio (defined by the ratio control) and at the same time increase the overall signal level.

 

This has the effect of increasing the overall loudness while not causing clipping.

 

Other things to look at woud be Gain Staging for track mixdown.

 

Understanding what dB is is useful when it comes to having a common language to talk to other recording engineers and producers.

 

Using your ears is of paramount importance, but knowledge in recording and production is essential when it comes to creating a quality recording of a predictable standard, and when it comes to being creative by intention instead of by happy accident.

 

I hope this helps!